Data & Analytics
Courses on working with data: querying, modelling, visualisation, and turning measurement into decisions.
Verified courses first, then the rest of the category
Forty hours covering data distributions, probability, sampling, hypothesis testing, inference and regression, aimed at business use. Free from Saylor, guest-readable, with a free certificate and college credit recommended.
Verification record: TMC-2026-00031Verified Re-check due Direct course review
Michigan’s applied data science sequence, one of the most taken Python data programmes published. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
The question-framing course in Google’s data analytics certificate, on deciding what to analyse before analysing it. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
How information systems are chosen and run inside an organisation, sitting between the business and technical courses in this catalog. Free from Saylor.
Not independently verified
An introduction to turning business data into decisions, covering the analytics stack at a conceptual level. Free from Saylor with a free certificate.
Not independently verified
Using data to decide rather than to report, aimed at managers rather than analysts. Free from Saylor with a free certificate.
Not independently verified
The advanced counterpart to Saylor’s introductory analytics course, for people building rather than commissioning analysis. Free from Saylor.
Not independently verified
How data is governed, stored and kept usable inside an organisation, which is usually the constraint on any analytics work. Free from Saylor.
Not independently verified
Presenting analysis so that it changes a decision, which is a distinct skill from producing it. Free from Saylor, readable as a guest without an account, with a free certificate on completion.
Not independently verified
Applying information systems to real organisational problems, following the introductory MIS course. Free from Saylor, readable as a guest without an account, with a free certificate on completion.
Not independently verified
Wharton’s business analytics specialization, covering customer, operations, people and accounting analytics. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
Modelling a business in spreadsheets and interpreting what the model says, from Wharton. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
Duke on choosing the metrics a company should actually run on, which is a harder question than measuring them. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
UC Davis on building dashboards in Tableau and making them say something. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
Machine learning taught as fundamentals rather than as a library tour, sitting alongside the statistics and Python courses in the same catalog. Free from Saylor, readable as a guest without an account, with a free certificate on completion.
Not independently verified
Python taught specifically for data work, following the introductory Python course and pairing with the statistics material. Free from Saylor, readable as a guest without an account, with a free certificate on completion.
Not independently verified
PwC’s own analysis and presentation method, which is unusual access to how a large consultancy actually works. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
Hypothesis testing taught as its own course, which is the step where most self-taught data work goes wrong. Free from The Open University, startable without signing up, with a free statement of participation on completion.
Not independently verified
Building data visualisations in Excel, aimed at the spreadsheet most analysis is still actually done in. Free from The Open University, startable without signing up, with a free statement of participation on completion.
Not independently verified
At one hundred and forty-seven hours, the longest programme on the DeepLearning.AI index. A beginner-level data analytics specialization rather than a short course.
Not independently verified
DeepLearning.AI’s foundations course for data analytics, sitting below its longer specialization. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
Five hours on pandas, the library most Python data analysis is actually done in, taught as interactive browser exercises. Requires a paid plan, with no price stated on the catalog listing.
Not independently verified
Illinois on data mining, covering pattern discovery and text mining at graduate level. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
The Johns Hopkins data science sequence, taught in R and among the longest-running programmes of its kind. Delivered through Coursera; the browse listing publishes no price.
Not independently verified
About this subject
Courses on working with data: the statistics underneath it, the languages and tools used to query and shape it, and the visualisation and communication that decide whether any of it changes a decision. The catalogue covers both ends of this, from spreadsheet fundamentals through to graduate statistics, and the gap between them is wider than course titles suggest.
Guidance
Verification tells you that a course delivers the material it advertises. It cannot tell you whether the methods it teaches suit your data. These are the things worth checking yourself.
A course on a business intelligence product teaches you to operate software. A course on statistics teaches you which questions the software can honestly answer. Analysts need both, and the second is the one people skip, because the first produces visible output faster. A dashboard built on a misunderstanding of sampling is worse than no dashboard.
SQL, Python, R, Excel and the major BI tools all appear in this catalogue and all lead somewhere different. If your workplace runs on spreadsheets, a course in R may be intellectually satisfying and practically irrelevant for a year. If you intend to move into engineering, the reverse. Each listing names the tool in the title or summary.
Hypothesis testing, sampling and regression are where most self-taught analysis goes wrong, because the tools will happily produce a confident-looking number from a method that does not apply. Several courses here teach statistics on its own, and taking one early tends to make everything after it faster.
Analysis that nobody acts on has not finished. The stronger courses in this subject cover how a finding is presented, what a chart encodes and how a recommendation is framed, not only how the calculation is done. It is worth checking whether a course stops at the output.
A course built on a clean teaching dataset is a different experience from one that makes you find, join and repair real data. Both have a place, but only the second prepares you for the part of the job that takes the most time. Where a provider describes exercises or projects, the summary says so.
Standards
The published standard applies unchanged. These are the checks that tend to matter most in this subject.
A verification says nothing about whether the statistical methods taught are appropriate for your data, or whether the conclusions in the teaching examples are sound. We do not review analytical correctness.
Questions
For a great deal of business analysis, yes. Most workplace analysis still happens in spreadsheets, and the courses here treat that as legitimate rather than as a stepping stone. Whether you also need SQL or Python depends on the size and source of your data.
Sometimes. Some courses run entirely in a browser, others assume a licence or a paid tier. Where a provider states what is required we record it, and where they do not, the listing says only what we could confirm.
That is between you and whoever is hiring. We verify that a certificate is described accurately, including its cost and what it requires, and we make no claim about its value in the market.
Related